A Novel Blind Source Separation Method Based on the Extended Natural Gradient

2012 ◽  
Vol 433-440 ◽  
pp. 3570-3576
Author(s):  
Yu Feng Xue ◽  
Yu Jia Wang ◽  
Qiu Dong Sun

In this paper, a new method is introduced to derive the extended natural gradient, which was proposed by Lewicki and Sejnowski in [1]. However, they made their derivation under many approximations, and the proof is also very complicated. To give a more rigors mathematical proof for this gradient, the Lie group invariance property is introduced which makes the proof much easier and straightforward. In addition, an iterative algorithm through Newton's method is also given to estimate the sources efficiently. The results of the experiments confirm the efficiency of the proposed method.

2011 ◽  
Vol 86 ◽  
pp. 180-183
Author(s):  
Yan Bin Lei ◽  
Zhi Gang Chen ◽  
Hai Ou Liu

A new blind source separation (BSS) algorithm used for separating mixed gearbox signals is proposed in this paper. Firstly, whiten the observed signals, and then diagonalize the second- and higher-order cumulant matrix to get an orthogonal separation matrix. The feasibility of the algorithm is validated through separating the mechanical simulation signals and the gearbox vibration signals. The algorithm can successfully identified the failure source of the gearbox and provides a new method to a gearbox fault.


Sign in / Sign up

Export Citation Format

Share Document